Lameness detection in cattle by using signal processing

Today, it is necessary to increase the productivity and profitability of dairy enterprises to meet the increasing demand for milk and to supply qualified food at affordable prices. Animal diseases should be combated in order to prolong the average life span of productive animals and to protect animal welfare in dairy farms. Foot diseases are usually seen after birth in dairy cows and can cause serious damage such as sending to slaughter. The common indicator of foot diseases is lameness. The treatment of lameness is a high-cost treatment. Apart from the treatment costs, loss of milk due to the use of antibiotics and loss of welfare of the animal should also be considered as harm. In addition, decreased fertility harms herd continuity. Combating against lameness can be done by providing early diagnosis of lameness after the environmental conditions are improved. Due to the increase in the number of animals in the enterprises, the time to observe the animals decreases and lameness cannot be detected during animal care. For this reason, there is a need for systems that can automatically detect lameness. In this thesis, an Artificial Intelligence Assisted Lameness Scoring System, which can be used to detect lameness in cattle and evaluate its severity, which is one of the leading problems of cattle breeders, has been developed. The system analyzes motion data using signal processing methods and scores lameness. Obtained results are shared with breeders and experts of the subject. Experiments carried out in the field show that the Artificial Intelligence Assisted Lameness Scoring System is successful and applicable.